Cardiac contractility: Correction strategies applied to telemetry data from a HESI-sponsored consortium

QT has a long history of heart rate (HR) correction but limited investigations have been undertaken to assess the impact of cardiovascular parameters on left ventricular (LV) contractility in drug safety testing. Cardiac contractility is affected by preload (Cyon-Frank-Starling law), afterload (Anrep effect) and HR (Bowditch effect). We evaluated multi-parameter correction methods to help with dP/dt_max interpretation.

Modeling was undertaken using data from dogs in single or double 4×4 Latin square studies. Correction models (16 fitting formulas×2 modeling approaches (universal and individualized)×2 correction approaches (linear or proportional)) were evaluated. 3D/2D cloud analysis of the beat-to-beat data for the control, pimobendan, and either itraconazole or atenolol groups were used to evaluate correlations between parameters and derive an optimal correction method.

Cardiac contractility (i.e., dP/dt_max) was best correlated to HR and systolic LV pressure with a correlation coefficient of 0.8. In decreasing order, dP/dt_min, mean arterial blood pressure (BP), systolic BP, diastolic BP, arterial pulse pressure and LV end diastolic pressure (LVEDP) showed a reduced correlation to dP/dt_max. Subject-specific models improved the correction by up to 14% when compared to universal correction models. The non-linear correction model was superior to the linear model.

Results suggest that the optimal correction formula for dP/dt_max would be subject-specific, non-linear and would include HR and LV systolic pressure. Correcting contractility for HR and systolic LV pressure may enhance data interpretation in non-clinical drug safety assessments. Similar correction methods could be evaluated for other species used in safety pharmacology.